Convert AVCHD to DV-AVI for editing?

Viewing 37 reply threads
  • Author
    • #44102

      I’m probably taking the plunge into an HD camcorder in the near future. However, I’m still using Premiere Pro 2.0 for editing, and upgrading _that_ and my computer system will cost a bit. So I’m wondering about converting AVCHD to DV-AVI so I can edit it (or try to edit it) with my current rig, until I spring for a new one. What are the tradeoffs– I mean, other than massive files on my hard drive? Is there a loss of color depth? Is DV-AVI limited to certain resolutions, framerates, color scales, etc.?

    • #184832


      Take Waiwai’s info with a handfull of salt. He’s just plugging product.

      Have you bought the camera yet? Are you in love with AVCHD or are you open to other HD formats? AVCHD is really compressed and every pro I’ve talked with about it says it’s a PIA to deal with.

    • #184833

      Why not Sony Vegas Pro? It supports AVCHD editing natively. Although it does not support smart rendering, it is okay for most editing works.

      If you only want to cut out unwanted scenes out the movie, you can try Smart Cutter for DV and DVB instead. Smart Cutter is aFrame AccurateH.264/AVCHD editor. Especially, only small parts on start and end points will be re-encoded, the middle part will remain untouched. This gives highest speed while reserving highest video quality. This is just ‘smart rendering’.

    • #184834

      Editing the AVCHD video natively is probably the best strategeom, though it may put a ton of lag on a non-core (e.g. Celeron) processor. If the speed handler is not a core, 2.5 GHz for editing AVCHD should probably be your minimum. Editing MPEG-2 footage on my duo-core 1.66 GHz processor isn’t bad at all with Sony Vegas Pro 9. The lag picked uptons more inCyberlink PowerDirector 7 (consumer video editing software). From my experience, the processing to lag ratio varies A TON among the software type. Pro quality software will generally edit more smoothly and produce less lag.

      Converting AVCHD to DV-AVI will limit the quality reproduction to the capabilities of both formats. For example, AVCHD and DV-AVI use different encoding systems (compression techniques and different types of losses). Converting highly compressed formats to DV-AVI will generally involve losses that are not noticeable enough to care. If the slow nature of AVCHD drags heavily, a DV-AVI conversion will probably be your best bet.

    • #184835

      “Is DV-AVI limited to certain resolutions”

      Yes…DV-AVI is standard definition (720×480, hence DV) so you’d essentially be converting an HD source to SD.

      There are other codecs that will allow you to convert to an AVI that will allow you to maintain HD quality like cineform’s neo scene ($129)

    • #184836

      edit native. once you start converting things left and right it gets’ pretty messy.

    • #184837

      “edit native. once you start converting things left and right it gets’ pretty messy.”

      There are basically two reasons one would convert AVCHD or HDV footage…Better performance within an NLE and better quality upon export of your final project….not sure what you mean by messy? It shouldn’t be a problem if you know what you’re doing.

    • #184838

      If at all possible, I’d agree about the part of editing natively, but a conversion between AVCHD to an HDV format would probably be (though I have no experience with it) virtually lossless. If the editing process becomes unbearable, or if you prefer a smoother workflow,then by all means I’d have to go with what Coreece said about converting.

    • #184839

      i think adobe cs4 can help u

    • #184840

      Adobe Premier Elements 7 can edit AVCHD native also.

    • #184841

      Premiere Elements 8 CAN theoretically edit native AVCHD files, but it is an acutely painful process even with a fast machine. And if you try to render and burn longer movies to DVD it will take many hours and may crash. I have a new quad core computer with 6 Gigs of RAM and a dedicated terabyte disc for video and the computer consistently crashes if the movie is longer than 40 minutes or so. Far better to convert your files to AVI format with a video converter program, such as AVS Video Converter, then edit in PRE. Judging from comments on various other forums this seems to be the way to go.

    • #184842

      All DV-AVI is in standard definition. Your best bet will probably be to convert the AVCHD files to a high bitrate MPEG-2 or H.264 file.

    • #184843

      You should check out the new Edius NEO2 w/ AVCHD Booster. It provices Real-Time AVCHD editing!

      Grass Valley EDIUS Neo 2 with AVCHD Booster Option
      The Perfect Solution for Editing AVCHD

      Native AVCHD files have previously been difficult to edit with NLE software,
      due to the high compression of the video, and often required dedicated
      hardware solutions. The new EDIUS Neo 2 Booster allows for smooth,
      full-frame editing of native AVCHD video in software only, courtesy of
      a newly-developed AVCHD codec engine that can process more than 3
      real-time streams simultaneously with EDIUS Neo 2 (tested with an Intel
      Core i7 CPU system – just like the Videoguys’ DIY 7 system configuration).

      GrassValley EDIUS Neo 2 with AVCHD Booster option is an affordable NLE
      software featuring basic, entry-level range of tools and features,
      EDIUS Neo 2 software is perfect for those new to video editing. Yet it
      also provides powerful nonlinear editing features common to all EDIUS
      solutions, such as real-time, multi-track, mixed-format HD/SD editing,
      chroma keying, titling, and Blu-ray Disc and DVD authoring. Built on
      the same highly-acclaimed technology as the EDIUS editing software,
      EDIUS Neo 2 offers native editing of various formats including AVCHD,
      HDV, DV, Windows Media, and QuickTime ?? and provides a seamless,
      real-time workflow that supports the mixing of all formats within the
      same timeline.

      The EDIUS Neo 2 Booster makes real-time AVCHD editing an affordable reality and
      enhances Grass Valley’s commitment to the real-time “anything in –
      anything out” editing platform. The EDIUS Neo 2 Booster is compatible
      with Windows XP, Vista (32-bit/64-bit) as well as 32-bit and 64-bit
      versions of the new Windows 7 Operating System.


    • #184844

      Try iOrgsoft AVCHD Converter.which can help u convert avchd to AVI/DV.

    • #184845

      why not try doremisoft video converter to convert your AVCHD to DV-AVI and also edit it (or
      try to edit it) with your files without loss of color depth.

    • #184846

      Well, here is what I do when I go to edit my AVCHD (.mts) files off my Canon HF S100 so that I can edit them in PP CS3. Iuse this.mts toXvid batch filethat was created by other people not some company and is free and easy.

      First I pull the files off my SDHC card into a preset folder

      Then I copy andpaste the batch and ffmpeg.exe into that folder and click on the batch file to transcode the video

      Then I import and edit 🙂

    • #184847

      If you are going to stick with your older computer system for a while then go with cineform Neoscene. You can get it for $99 and it will allow you to convert to it’s intermediate codec for ease of editing, then you can go directly to burn an HD Bluray disc, or go back to AVCHD w/o any loss of quality. It will also give you a much wider choice of NLE’s when you do upgrade because you won’t have to worry about lag. Get Cineform and be worry free


    • #184848

      I use New Blue FX’s AVCHD Upshift – It converts MTS files to lightly compressed M2T files (MPEG-2) which I then edit in Vegas nicely on an older Pentium IV running at 2.53 GHz with 1 GB RAM (it works better on a faster box with more RAM).

    • #184849

      I use total video converter converting first.

      E.M. Total Video Converter is a piece of
      extremely powerful and full-featured converter
      software that supports almost all video and
      audio formats. The software is designed to
      convert video for your mobile video player as 3gp,
      mp4, PSP, PS3, iPod, iPhone etc and also VCD or DVD player,

    • #184850

      so, i have purchased the HMC150, downloaded the panasonic recomended transcoder and started doing some testing, Great camera!

      now, i have noticed that when i have converted the file to a DV-AVI 16:9 video, i noticed that the video is fuzzier and a few pexils wider! for all of you who need to check this out, open your choice of NLE (i use vegas pro 9), drop the .mts video and the transcoded videos on the timeline and set up multicam edit or press play to see the video, you will see the difference in the video.

      i want to properly transcode the video to keep the same video quality, just as a DV-AVI for easy editing with another camera. please check out to see what i am talking about, and help me figure this out!

    • #184851

      are there any free AVCHD converters for macs?

    • #184852

      DV-AVI is generally the format with less compression and quality loss. The way I am using to conert AVCHD video form my HD camcorder is using Aunsoft MTS Converter for Mac which is the easiest and fastest program I have ever used to accomplish AVCHD video conversion.

      Step 1. ImportAVCHD files to this program.

      Step 2. Choose DV AVI format.

      Step 3. Convert VRO to DV AVI for Mac.

      Quite easy to use.

      And the most satisfied one is its high quality of output video formats.

    • #184853

      I’ve had good results converting the .mts files to .mov, with the settings h.264, 12000kbps, 1280*720, 25fps, aac. The files look good on my Mac running Adobe Premiere, edit easily, and convert well to DVD format.

    • #184854

      If I might introduce myself as a dyed-in-the-wool PC Adobe Premiere user, CS5 to be exact, I can attest that CS5 does an okay job with native AVCHD, but truthfully, I have not yet tried long form video of an hour or more. Computer is a quad AMD with 8-G of RAM, 3-GHz processing with Blackmagic Intensity Pro, and 64 bit. (No Mercury Engine).

      I am reading that when the computer begins to choke, which appears to happen around 40-min. or so, I do wish to consider a conversion, if that will help. If H.264 is best, or if DV-AVI can be done to retain the 16 x 9 A/R, does anyone know which might be best to minimize image quality loss or provide best editing, or both ? There are many conversion options built into Premiere’s Encoder, but no one suggests which is best for any particular circumstance. A lot of my long form video occurs in Wedding work, where I can easily have 2-hrs or more of raw footage.

      HD Camera, BTW, is a Sony HDR-SR-12, a real sweetheart single CMOS unit with a 120-G Hard Drive, and when 1080i footage is shot at FS speed, (a higher speed than the normal SP), the image, under good lighting, is completely comparable to what is put out with much more expensive and very high end HD camcorders.

      For me, one of the big treats in using this camcorder is that the AVCHD files, using a USB cord connected to the camcorder, can copy and paste video files directly into the computer without having to “capture” the footage at real time, which I have always had to do with my PD-150 SD-DV tape camcorder.

      If, however, to facilitate editing, I am advised to convert this footage to something else, I may need to trade the bullet of saved time for easier editing in another format after conversion.

      What would seem to be the best way to proceed ?

    • #184855

      Hi Robert –

      I too shoot with a SR11 (same camera, smaller hard drive). I have had good luck with New Blue FX’s UpShift (see comment above). It converts MTS (AVCHD) to M2T (lightly compressed MPEG-2) which I edit easily in Sony Vegas. It can be done in a batch mode, so I leave large conversions running overnight.

      FWIW – I loves me my SR11 (wish I had sprung for the extra HD space of the SR12) as it has many features of a true prosumer camera (manual focus & exposure, Dolby 5.1, headphone jack, mic jack, optical image stabilization, HDMI out, 1/3″ CMOS sensor, Zebra stripes, manual & auto white balance, HD recording @ 1080i, large viewscreen, etc…) for a (high end of) consumer price.

    • #184856

      Hey Birdcat….thanks for the reinforcement of the many attributes of that camcorder design, too easily taken for granted ! Not being where I should be in the vernacular department, what is the “batch mode” you describe ? …. a partially processed timeline ? And when you convert to M2T, do you retain the 16 x 9 A/R, (or 1.0 pixel ratio?) I’m not sure if M2T is an option in Encoder ….. will need to check. And for that conversion, does the timeline content then edit any easier than if you leave it in AVCHD (MTS) mode ?

      Another issue I did not discuss is the slight nuisance of the SR-12 recording HD in 2-GB segments or clips. This is noticeable in long form video recordings, wherein a gap is produced after each 2-Gs of recording, disrupting desired continuity (without gaps). Coming to the rescue for this, I understand that an “MTS Merge” program is supposed to connect these clips and eliminate any discernible gap. Are you familiar with that ? I have yet to get that into motion, but something like that is going to be necessary if I am to combine SR-12 footage with a second camera, of some genre, and not miss a beat in sync.

    • #184857

      I edit AVCHD with no problem in PP CS4. I would imagine with your system you would have a problem. The most reliable converter is NeoScene. It is not free but it really works. You can find it here with new software and fast computer it will make your life easier if you convert.


    • #184858

      Hi Robert –

      You are correct in the 2GB gap – I usually pause before I reach that limit but it has bitten me before. I haven’t heard of anything to solve this but it hasn’t been high on my list as I try to keep clips 10 minutes or less (easier editing).

      As for batch mode, UpShift allows you to specify a directory, and will convert all files therein, sequentially, until they’re all done – I’ve done 15-20GB at a clip when needed, albeit it took about a day on my box.

      And yes, the M2T files are also 16:9, square pixel. Sony Vegas Pro reads the M2T files natively (it is lightly compressed MPEG2). If you have an issue with M2T, you can also look at GearShift from VASST which does something similar but outputs various formats.

    • #184859

      You mean your video editing can’t occupy more than 2GB of RAM?

    • #184860

      First to XTR-91: No… during recording, say of a 2-hr event, the AVCHD SR-11/12 records in only 2GB segments, and then produces a gap of a few seconds to sort of mess up continuity before continuing its recording operation. The camcorder was never turned off or lets you know this is happening….it just records in segments, which you discover when you place the clips on the timeline to edit.

      Now for Birdcat : The good news is I just found the lost application that solves this problem of joining the segments. It is called …. MTSMERGE.exe A great way to input all the 2-GB segments and the FREE app nicely links them together into one clip, which of course in shooting, was the intention all along. So that would seem to solve the segmentation problem. I plan on getting into editing that long one tomorrow, and I’ll get back if I discover any problems. BTW, some HD camcorders do this segmentation-recording in 4-GB segs, but either way, when doing long form video editing, that needs to be continuous, that little habit is a real nuisance. So merging the clips is a real necessity.

      But back to the M2T file conversion, I really need to see if that conversion codec is available in CS5 Encoder, because every filter or cosmetic correction I add now to the MTS video stream takes “forever” to render, compared to good ole’ DV. So much complexity here, that it is difficult to even scrub thru the clip before rendering. Hence, if a different file is going to be easier to edit, and we can still retain original resolution and clarity, etc. – that should be a good way to go.

    • #184861

      Addendum/Correction to Birdcat : File is actually MTSMERGER.exe

    • #184862

      Hi Robert –

      I guess my question would be where does it get the missing audio.

      I can crossfade in my NLE so the sound appears contiguous (although that is not optimal) or live with the approx half second audio gap – As I said, I usually pause every so often so I don’t run up against this often.

      I will look up the MTSMERGER though – Thanks for the tip.

    • #184863


      On my HMC150, if I import the entire folder of AVCHD files, it should import the metadata as well which includes MXF files (and tells where to span the video files from one to the other). When I do this, I have no gap in Audio or Video. However, if I simply import the clips from the clip folder, I have to merge them separately and do experience a little gap in the Audio and Video.

      One thing I think is important to remember about AVCHD and it being compressed is that Color tweaking in Post is more limited than uncompressed footage. Those who are used to the feature coming from DV SD footage may be sorely disappointed. When you shoot with AVCHD, it is ultra important to get the shot right.

      In terms of the Adobe suite, I started with CS4 shortly after purchasing my HMC150 a little over two years ago. version 4 was glitchy when working with AVCHD natively and I often experienced freezes and hangups.. However, when 4.1 update was released, most of these problems went away for me. I was sorely disappointed with the performance of 4.0 and even made calls to my camera vendor (Who in turn put me in touch with the Panasonic rep who in turn put me in contact with a developer from Adobe). He is the one who pushed me to do the update to 4.1 and I was absolutely ecstatic with the results. Granted, I was using a Core I7 920 with 6 Gig of RAM at the time.

      When CS5 came out, I took the plunge and upgraded. I also upgraded to 12 Gig of RAM as well as got myself a Nvidia GTX285. The results for native editing were indeed astounding. I am able to apply many transitions and filters and am able to preview them without having to render first and this saves me much time. However, when it comes time to transcode to a different format for output (Through Media Encoder), This is where time is not saved, It still has to decompress the AVCHD footage and in turn, causes longer transcoding times. Because of the nature of GPU processing (it could be described as “Quick and dirty” in terms of video rendering), it works for preview but actual transcoding is done by the computer processor.

      Still, with mercury, time is saved by not having to transcode to edit easily. It truly is convenient and though I still have long transcoding at the end, It really does make my workflow easier because I don’t have as many “Intermediary transcoding” when applying filters or transitions to see how they look. not only, Mercury makes it much easier to do PIP editing with AVCHD because it pretty much emulates the look on the fly (Again without having to render an edit to see how it looks).

      I also work with Final Cut Pro at work and though all AVCHD is logged and transferred into the Pro Res medium, it seems to be a much quicker procedure of transcoding with less processing power. The only disadvantage is that the transcoded footage is about 3 to 4 times larger in file size than the AVCHD footage. Maintaining a larger data drive will be important for project performance and management.

      All in all, one thing I often see overlooked in terms of Video editing performance on a PC is “dealing with Fragmentation”. Before I do a long transcode, I always defragment the drive first. Severe fragmentation from copy large files here and there severely slows performance when it comes to transcoding large files. To ensure quick transcoding, I always suggest defragmenting before you perform the transcode (or even render).

      I did, as a precautionary measure, purchase Cineform and sometimes do use it (especially when combining footage from my HMC150, HMC40 and my 5d MK II). It just keeps everything in the same preset workflow and makes things much smoother. Other times, I will just simply edit the AVCHD footage natively.

      Hope this wasn’t too much for my first post!



    • #184864

      Hi Anton –

      Long answers are very good as they contain much needed detail.

      On my SR11 I do not have MXF files (there are CPI files which I believe contain the clip information) and the raw MTS files (or MPG if recording in SD).

      What I was speaking of concerning the missing audio is a glitch (the only one I’ve run across) in the design of these cameras (SR11/SR12) that Robert pointed out – When recording longer segments, the camera breaks the video into maximum 2GB files and has a very small gap in the audio between each.

      I am unaware of any files on the camera which contain this missing audio and, as in Robert’s case, I was curious where the MTSMERGER program acquired this missing data from.

      As for the MXF files, I was under the impression that Panasonic’s flavor differs from others – I know there are NLE’s that cannot read them natively (look here for a solution to that problem:,

      Concerning the color tweaking – I have not noticed significant differences in my NLE (Sony Vegas Pro 8) in color correcting between the HD M2T files (output from UpShift’s MTS conversion) and my old captured SD AVI files. (But that could just be these old eyes…).

    • #184865

      Although I feel a bit of a traitor, since it has served me well, I am at last moving beyond the 720 x 576 format, (in case you wondered, I am talking New Zealand and PAL_B here), and I have been looking for a means of modernising my setup. Unfortunately, I amcaught more than halfway through a long-term undertaking, a documentary series covering the entire coastline of the province (of NZ), where I live, now into its sixth year and 96th DVD of source material. Being ‘far from civilisation’, and having, until recently, belonged to a club which thinks video is going to meander along forever based upon the4 x 3 aspect-ratio and the memers’ ‘steam-powered’ methods of achieving it’, I have had to do a lot of lateral thinking to secure the results I want; eg an easy pathway to having everything, (good quality standard footage included), upscaled to 1920 x 1080. Here is the outcome of some of my ‘lateral thinking’ on the subject.

      Beginning with the premise that the outcome should determine the method, where possible, I have settled upon non-standard ‘custom’ format of my own, 960 x 540. Does that ring a bell? Linearly, it is exactly half the pixel-count in both directions, of 1920 x 1080, my desired outcome. Each square pixel, then, multiplies itself by four, making the easiest possible (if rather slow), going for ‘h264’ (the programme), which does the ‘grunty’ part of the upscaling. To create a 960 image, puts the camera output, (which has been compressed into 720 x 576), back to where it started out. I arrived at that figure both by calculation, and my experiences from having to adapt some still graphic images to fit seamlessly into an editor’stimeline, with the minimum of fuss, as they, of course, had not been laterally compressed as had been the video images. 960 x 540 was where they came-out correctly with the aspect-ratio intact, although many dimensions of the correct ‘x to y’ ratio will also work.

      First-up the image is adjusted slightly, it has 18px lopped off top and bottom, to bring it to the exact 16:9 aspect-ratio and the correct dimensions to be an easy upscale. I have used ‘Virtualdub’ for both of these functions, eg the cropping, and the re-scaling.

      The problem then arises as to when to best do the cropping and processing. I feel it makes good sense to do at least ‘chapters’ and if the features are not overly long, say sevento 20 minutes, which is what my production breaks down into in a ‘modular’ fashion, possibly doing the complete conversion process on the finished product.If need be, the video/audio streams may be separated for this, and re-multiplexed afterwards. By my reckoning it makes very good sense to do it that way and so, instead of prematurely transcoding shots tompg2 as I used to do, it pays to keep footage as DV AVI as long as possible, to keep ‘Virtualdub’ happy. Because most editing software, (at least at the level I am able to afford), goes ‘tits-up’ when confronted by DIY formats, it is best, in my experience, not to antagonise it by calling upon it to edit and render images in 960 x 540, however, as a final encoding, free of the restraints of the timeline, there should beno problem. If all else fails, it is likely TMPGEnc will be able to help with re-encoding using ‘custom’ settings. (For those of aninquiring frame-of-mind, AVS editing software from Britain, does at least accept 960x 540 for editing, but my experiments have not yet gone beyond that).

      The techniques I have described are well into the advanced stages of being proven and my only worry is that ‘Virtualdub’ might ‘choke’ and fall over on really ambitious processes, withconsequent crashes. So, experiments continue.

      The time taken to process, of course, will make ‘over-nighters’ almost inevitable rendering-wise, so it is best to simply go off-to-bed and leave the computer to-it.

      If you wonder why I went this way about the process? I had been brought up on ‘widescreen’, from my film days and anamorphic lenses. I foolishly started my documentary series in 4 x 3 aspect-ratio to keep in uniformity with other members of the club I belonged to. That left me with some hours of footage which had to be cropped back to ‘widescreen’ dimensions using ‘GIMP’, which for all that I love it to bits, is a bit of a grisly thing to have to contemplate when it comes to dealing with some hours of footage. The method I have described, grew out of a lot of experience of having ‘widescreened’ 4 x 3 and is an adaptation of a process I have carried out many times, only ‘Virtualdub’ makes it a whole lot easier.

      Ian Smith – Dunedin, New Zealand

    • #184866

      ‘Kate59588’: Just to update a previous posting, I have been carrying out a similar process to the one you mentioned, using h264, which gives small files and very good results. The original footage was in 16:9 aspect-ratio, squeezed to 720 x 576px, as is usual. Using ‘Virtualdub’, I uncompressed the footage to a format of 1024 x 576, as I understand that is what my camcorder ‘saw’ originally. To eliminate any losses I set the ‘sharpness’ setting very low (eg ’22’) and recorded the stuff back to AVI type-2.

      From there, TMPGEnc ‘Video Mastering Studio 5’ did all subsequent processing, upscaling to 1280 x 720 in x264 with aac audio. The original footage in DV AVI was pretty good to start withand seems to have lost little in the upscaling. The tad of extra sharpness, in fact, has added noticeable detail which originally ‘wasn’t there’. I am not kidding myself that what I have produced is ‘High Definition’; I know it isn’t, but it’s damned good ‘SD’ for all that.By the way, why I used ‘Virtualdub’, (which might seem a bit on the ‘steampowered’ side), is because it has an excellent monitor which allows side-by-side viewing of the outcomes. There you are actually ABLE to see the extent of any improvements; not so, with most monitors.

    • #208719

      Aunsoft Video Covnerter for Mac can covnert AVCHD to DV-AVI with high quality for editing.


    • #214365

      Just try an easy-to-use AVCHD MTS converter(
      I am very pleased with it’s performance and capabilities and it is exactly what I was looking for.
      What I really like about the product is that it allows me to edit them down before converting them which really simplifies everything and frees up lots of space on both the pvr hard disk.

Viewing 37 reply threads
  • The forum ‘Video and Film Discussion’ is closed to new topics and replies.


Panasonic LUMIX GH5s w/ Leica lens

Given away to one lucky winner

Competition is open worldwide