@EricLNZ Thank you

I've been quite happy with the results of Mercalli 6 OFX using the trial version in Vegas Pro 23, so I've just used the plastic-fantastic to purchase it today. It's still 30% off but when that offer ends isn't given on the ProDad website.

Yes, I confirm it too. I purchased Mercalli 6 OFX few weeks ago and it is good with stabilizing footage. I stopped using Mercalli 6 SAL.

@Dexcon

I have been wondering if Mercalli 6 OFX is faster with analyzing and during rendering when it is applied compared with Mercalli 5 OFX or 4 OFX.

I have been wondering if Mercalli 6 OFX is faster with analyzing and during rendering when it is applied compared with Mercalli 5 OFX or 4 OFX.

I'm not sure mainly because the era in which I used Merc 4 for Vegas a lot was on a much older computer and with much earlier versions of Vegas Pro. I did try Merc 5 for Vegas as a trial but found no difference between it and Merc 4 for Vegas (keeping in mind that the leap from Merc 4 SAL to Merc 5 SAL was free because there were no new features, just technical changes).

BTW, I've just rendered a short length of a 9 mins project with lots of other FX (inc BCC Flicker Fixer) along with Merc 6 OFX, but I had loop selected in render options so only about 30 secs got rendered. But the good news is that stablisation using Merc 6 OFX worked well.

From all stabilization tools I tested and compared (Vegas stabilization, Mercalli 4 plugin, Mercalli 6 OFX trial, Mercalli 5 & 6 SAL) they were easily beaten in analyzing speed and result by the stabilization of Davinci Resolve (free). I wish Vegas would have such an easy to use, fast and high quality stabilizer built in.

@3POINT  ... In general, I agree that Resolve stab is excellent (especially with speed) but using the same footage I've very recently used with Merc 6 OFX in VP, Merc 6 OFX provided better stability on a quite slow drive up a tree-lined boulevard (historic 35mm film video I should add). Resolve's stab unfortunately resulting in some 'bending' of the tree trunks - which was not evident with Merc 6 OFX and totally unexpected in Resolve. Quite surprising IMO given the excellent results (but not always) that I usually get from Resolve stab.

@Dexcon, does Mercalli 6 OFX show up in Resolve?

@Dexcon, does Mercalli 6 OFX show up in Resolve?

Yes, but is much slower as DR stabi, result is the same I would say at first sight..

Thanks. I've yet to see it there. I'll download and try again. I am just curious, that's all. Again, thanks!

I am confused and annoyed. For me, unfortunatelly, Prodad Mercalli OFX v6 is useless :(((((((

I tried to stabilize the shaky clip. It is 4K 200Mbps 50p 4:2:2 10 bits (SLOG3) video on my PC.

I know that Vegas Pro 23 has not been still supporting this kind of file with my GPU Nvidia RTX 5090, that's why I was forced to

* turn on in Preferences-> File I/O-> Enable legacy AVC decoding and

* turning off Hardware decoder to use: (I have Off setting).

Video was analyzed by Prodad Mercalli OFX 6 sucessfully, but unfortunately only in preview in Vegas Pro 23 it is not shaky. When I render video to Voukoder Pro to AV1 codec it is still shaky in output mp4 file.

I share with you a sample. Maybe this sample is too short for Prodad to stabilize, but if it was to short, the preview window would display shakeness, but in preview in Vegas playback is fine without shaking.

Maybe the most important thing I did was the fact that I shorten the event to 1 sec 160 ms. The original event was 2 sec 880 ms.

This 1 sec 160 ms plays without shaking only in preview window, after rendered it is still shaky.

This 2 sec 880 ms event (original media length) is good both in preview and in output file.

Both events were processed with ProDAD Mercalli OFX 6.

Here is the file:

https://marcinzmyslowski.synology.me:444/download.php?plik=Sony%20FX3_oryginal_2026-02-06%20-%202026-02-06%20-%20Gdansk%20448.MP4

Can you help me finding workaround for stabilization this file?
I don't know what to do, because mostly I edit very dynamic videoclips and that's why I have such short events.

I can even write that output rendered file with Prodad stabilization OFX 6 applied on 1 sec 160 ms is even worse than original clip with the same shorten length rendered to output file with no apply of Prodad stabilization OFX 6.

I also see that it has no matter if I render to Voukoder, Magix HEVC, mpeg, Magix AV1 codecs, becuase the shakeness is the same in all of these codecs rendered.

Unfortunately I cannot confirm your issue, when I shorten your clip to 1 sec on a VP23 timeline and stabilize with M6 OFX (Demo), the rendered 1 sec clip is also stabilized.

1 sec clip unstabilized and the the same clip stabilized:

When you want such short stabilized clips I would think about buying a photo camera...😂

As you have pointed out, the original video is 50 fps. Using Mercalli OFX 6 and using an AVC (H.264) UHD 50 fps render template, the Merc 6 stabilisation worked well in the render. However, when using an AVC (H.264) UHD 25 fps render template, the Merc 6 stabilisation sort of worked but was very jittery. Transcoding the 50 fps original to 25 fps and then applying Merc 6 to that 25 fps render did not improve things - it was actually worse than before. Transcoding with Optical Flow activated didn't help.

What did work, however, was to take the successfully stabilised 50 fps render and re-render it to AVC (H.264) UHD 25 fps - the new 25 fps render maintained the stabilisation. EDIT ... if you are in fact wanting to render to 25 fps rather than 50 fps.

Why should @marcinzm record 50p and than render to 25p?

@marcinzm  ... Were your renders with Merc 6 to 50 fps or to 25 fps?

@marcinzm I do not use HEVC, nor have 4:2:2 10 bits source video, nor use 50P, I use 4K 25P from Sony RX10 to FHD projects in Vegas. I may have mentioned this before:
The fact you have to use file/IO decoding set to enable legacy AVC decoding in VP23, just like me (I have also a NVIDEA RTX 5090) is perhaps due to the 5090 or to a different "enhanced" HW decoding treatment or pipeline in VP23, that new decoding is not without it issues, in fact a project in VP23 with stabilization done in Mercalli with the file/IO decoding set to enable legacy AVC decoding in VP23, will have to redo the analysis if the legacy setting is back to its default (disabled).
So even when staying in VP23 you may need to re-analyze again depending on the decoding setting.
Probably the Mercalli stabilizing is not broken from VP22 to VP23 if the VP22 project in VP23 is opened with legacy AVC decoding enabled.

Now coming back to 50P to 25P and Mercalli OFX, I wonder on where you place the Mercalli OFX in the OFX chain in your project and how that may affect the functionality: by default placing FX are put after the pan/crop, but placing the Mercalli OFX before the pan/crop will (as I assume, I could be wrong) treat the original source, and not converted by pan/crop (for example FHD project if source is 4K). I wonder if Mercalli OFX is working on 25P or 50P depending on it place in the chain, if it matters, you will have different results for sure.
You may try to place the Mercalli OFX on media level rather than event level...

When it's about rendering 50p to 25p, just start your project with a 25p project setting.

Mercalli can't be used as media FX. I would always use Mercalli on (trimmed) events level and but it before Pan/crop.

When it's about rendering 50p to 25p, just start your project with a 25p project setting.

A very good point given that my earlier testing was with 50 fps set in project properties. Changing properties to 25 fps did improve the 25 fps stabilisation render but with some unacceptable jerkiness. There was no obvious difference with stabilisation smoothness with the Merc 6 plugin placed before or after pan/crop.

So far, the smoothest stab result I've achieved with a 25 fps render was the one where it was first rendered to 50 fps and then that render re-rendered to 25 fps.

Just as a matter of interest, I tried the 2 x Merc 4 for Vegas plugins (CMOS and Stab) and got the same result as with Merc 6 with the 50 fps properties setting and 25 fps render: i.e. it didn't work.

Vegas Pro 23 b302.

Please keep in mind that I do not have (as yet) a 50xx series NVIDIA GPU - I'm still on a 2080 Super GPU so my results may well differ from those of others.

@Dexcon I don't understand why you bring up 50p to 25p rendering, I assume when @marcinzm records in 50p, he also delivers in 50p. Also when you do not start a project with 25p, but in the middle of the project reset to 25p, than analyzing with Mercalli has to be redone.

Also I just have an ancient GTX2080, which works fine and I do not need to enable legacy AVC encoding.

@3POINT  ... let's wait until @marcinzm responds about what fps template he is rendering to - but I got the same result as what he experienced when I did the 50 to 25 render (as a reminder, 50 to 50 worked fine).

Also when you do not start a project with 25p, but in the middle of the project reset to 25p, than analyzing with Mercalli has to be redone.

It was re-analysed - every change in every test I made was re-analysed - including before/after pan/crop.

Also I just have an ancient GTX2080, which works fine and I do not need to enable legacy AVC encoding.

Legacy AVC is not enabled in File I/O in my VP23 - it was bitman who mentioned that issue.

I always record videos in 50p and render it to 25p. Some of my videoclips 50p used in project are slowed down to 0.5x so it has 25p but final render is always 25p.

@marcinzm So you throw half away of what you recorded and loose smooth motion? I hope you record at least with 180 degree shutter to get appropriate motion blur.

I would deliver in 50p and for those slowmotion clips you can just stretch them to 25p (showing the 50p frames double) or use optical flow resampling for real smooth results.