AJA T-Tap vs Blackmagic Mini Monitor test

[MIKEROPHONICS]

Dear All

I have done a comparison test on Pro Tools Expert between these two devices.
Due out later today
Hard hats advised....

[Frank Kruse]

Mike,

not sure what to make of this comparison. You used a different OS and different PT version and even a different Computer.
The past has shown that any of these factors can change the video behavior a lot.

The only thing I can read of this test is that the BM-box is useless in that combination of OS, PT version and driver also when used on a "Mac PC". What does that mean BTW? Did you use a Mackintosh (PC hacked to run OS X)? Because that could be another factor.

A colleague next door is using that exact box with a 6-core trashcan without the massive span you measured. Maybe because he's using a different CPU or BM driver. He's also using mainly 1080P/24.00fps films.

Also may I ask when you write about "PAL frames" using a 25P video with the output set to 50i but using the 25fps setting on the syncheck, what does "one PAL frame" refer to ?

Have you tested what happens when you use the correct 25P output setting for your 25P video? Just wondering if that makes a difference.

You used the 10-bit setting and the draft setting. Did you try and use the "full" regular (non 10bit) setting that I assume most people use?

You also wrote "no video references are required" which I think should rather translate to "you can not gen-lock the AAJ or the BM boxes because it doesn't offer a video-ref input which doesn't allow for frame-edge accurate sync lock with Protools"


Just trying to understand what you did because I didn't see these figures last time I checked the MB box in our studio. I would have returned it to be honest had it performed that way.

Frank.

[paulo m]

Originally Posted by Frank Kruse

Mike,

not sure what to make of this comparison. You used a different OS and different PT version and even a different Computer.
The past has shown that any of these factors can change the video behavior a lot.

The only thing I can read of this test is that the BM-box is useless in that combination of OS, PT version and driver also when used on a "Mac PC". What does that mean BTW? Did you use a Mackintosh (PC hacked to run OS X)? Because that could be another factor.

A colleague next door is using that exact box with a 6-core trashcan without the massive span you measured. Maybe because he's using a different CPU or BM driver. He's also using mainly 1080P/24.00fps films.

Also may I ask when you write about "PAL frames" using a 25P video with the output set to 50i but using the 25fps setting on the syncheck, what does "one PAL frame" refer to ?

Have you tested what happens when you use the correct 25P output setting for your 25P video? Just wondering if that makes a difference.

You used the 10-bit setting and the draft setting. Did you try and use the "full" regular (non 10bit) setting that I assume most people use?

You also wrote "no video references are required" which I think should rather translate to "you can not gen-lock the AAJ or the BM boxes because it doesn't offer a video-ref input which doesn't allow for frame-edge accurate sync lock with Protools"


Just trying to understand what you did because I didn't see these figures last time I checked the MB box in our studio. I would have returned it to be honest had it performed that way.

Frank.

Couldn´t agree more Frank.

[tom_lowe]

Originally Posted by Frank Kruse

Mike,

not sure what to make of this comparison. You used a different OS and different PT version and even a different Computer.
The past has shown that any of these factors can change the video behavior a lot.

The only thing I can read of this test is that the BM-box is useless in that combination of OS, PT version and driver also when used on a "Mac PC". What does that mean BTW? Did you use a Mackintosh (PC hacked to run OS X)? Because that could be another factor.

A colleague next door is using that exact box with a 6-core trashcan without the massive span you measured. Maybe because he's using a different CPU or BM driver. He's also using mainly 1080P/24.00fps films.

Also may I ask when you write about "PAL frames" using a 25P video with the output set to 50i but using the 25fps setting on the syncheck, what does "one PAL frame" refer to ?

Have you tested what happens when you use the correct 25P output setting for your 25P video? Just wondering if that makes a difference.

You used the 10-bit setting and the draft setting. Did you try and use the "full" regular (non 10bit) setting that I assume most people use?

You also wrote "no video references are required" which I think should rather translate to "you can not gen-lock the AAJ or the BM boxes because it doesn't offer a video-ref input which doesn't allow for frame-edge accurate sync lock with Protools"


Just trying to understand what you did because I didn't see these figures last time I checked the MB box in our studio. I would have returned it to be honest had it performed that way.

Frank.

Hi Frank,

I believe Mike tested the BM Mini-monitor on Alan Sallabank's PC as well as his own Pro Tools PC (running OS X) which was the exact same system he used to test the AJA T-tap. Seems a pretty fair comparison to me.

As for the lack of Ref input, the results from the AJA T-tap show this is no where near as necessary these days as it was in the past. The only real reason I can see for Ref input is if playing out to video decks.

[MIKEROPHONICS]

I have abandoned my BM Decklink Studio card and gone to an AJA IOXT as my previous experience showed me that BM drivers are just plain unreliable at best.

During the time that I was having endless problems in PT11 with my BM hardware. Alan was having far less grief on the windows OS. Ok it was a different card to mine, but my ears pricked up.

Having AJA hardware compared to BM is night and day in my experience.

When the T-Tap came out I tested it to see how good an Un referenced card can be - stonkingly good is the answer. With the difference barely worth the cost.

I wanted to test the BM equivalent and try a comparison to windows as best as I can.

I am not a test lab with unlimited resources working under nitrogen atmospheres and have limited time as a busy post pro, but I make an effort to publish my findings to help others.

I started with two hypothesis;
1) BM drivers are better on windows based on anecdotal evidence from Alan
2) The Min Monitor should give reasonable non referenced results, based upon my experience with the T-Tap.
3) BM do not write good drivers and playing driver roulette is unavoidable.

The results - yes slightly limited and slightly flawed in the lack of 100 % comparison don't disprove 1), disprove 2) and seem to agree with 3). Take from them what you will.

Any tests that are not on an identical rig to yours will not be ideal for you - for that I can't help.

It Feinstein points to The latest BM driver being as bad as many of the others, which is why I bailed on them. The latest AJA driver however is absolutely fine on 10.10.3 and worked out of the box on my IOXT. Chalk and cheese.

I have not found any software that cares about my hackingtosh. It is very close to mac hardware and is chosen for the task by experts who I have confidence in. So I discount that.

The BM would not work without flashing on 10 bit full, (which is my default setting on AJA hardware) or the intermediate setting, but would only work on the draught setting.

I rarely use 24f video as I oredominately lurk in TV, so tested 25 frame. The file allows PT to output 25p or 50I from the same file with the same reference. Identical results.

I am sure the BM might work somewhere with some OS/driver settings, but for me I demonstrated that the latest BM driver is duff. This compared with my other experience makes them a poor choice imho, with maybe windows users faring better. Yours may be working for you, which is nice for you, but....

I am not trying to offend anyone, just sharing my experiences with the community and trying to make any tests as best as I can within limited time and budget.

Please feel free to conduct your own tests and publish them and we will all read them with great interest (when you have some spare time of course).

Have a nice weekend.

[tom_lowe]

Originally Posted by MIKEROPHONICS

I have abandoned my BM Decklink Studio card and gone to an AJA IOXT as my previous experience showed me that BM drivers are just plain unreliable at best.

During the time that I was having endless problems in PT11 with my BM hardware. Alan was having far less grief on the windows OS. Ok it was a different card to mine, but my ears pricked up.

Having AJA hardware compared to BM is night and day in my experience.

When the T-Tap came out I tested it to see how good an Un referenced card can be - stonkingly good is the answer. With the difference barely worth the cost.

I wanted to test the BM equivalent and try a comparison to windows as best as I can.

I am not a test lab with unlimited resources working under nitrogen atmospheres and have limited time as a busy post pro, but I make an effort to publish my findings to help others.

I started with two hypothesis;
1) BM drivers are better on windows based on anecdotal evidence from Alan
2) The Min Monitor should give reasonable non referenced results, based upon my experience with the T-Tap.
3) BM do not write good drivers and playing driver roulette is unavoidable.

The results - yes slightly limited and slightly flawed in the lack of 100 % comparison don't disprove 1), disprove 2) and seem to agree with 3). Take from them what you will.

Any tests that are not on an identical rig to yours will not be ideal for you - for that I can't help.

It Feinstein points to The latest BM driver being as bad as many of the others, which is why I bailed on them. The latest AJA driver however is absolutely fine on 10.10.3 and worked out of the box on my IOXT. Chalk and cheese.

I have not found any software that cares about my hackingtosh. It is very close to mac hardware and is chosen for the task by experts who I have confidence in. So I discount that.

The BM would not work without flashing on 10 bit full, (which is my default setting on AJA hardware) or the intermediate setting, but would only work on the draught setting.

I rarely use 24f video as I oredominately lurk in TV, so tested 25 frame. The file allows PT to output 25p or 50I from the same file with the same reference. Identical results.

I am sure the BM might work somewhere with some OS/driver settings, but for me I demonstrated that the latest BM driver is duff. This compared with my other experience makes them a poor choice imho, with maybe windows users faring better. Yours may be working for you, which is nice for you, but....

I am not trying to offend anyone, just sharing my experiences with the community and trying to make any tests as best as I can within limited time and budget.

Please feel free to conduct your own tests and publish them and we will all read them with great interest (when you have some spare time of course).

Have a nice weekend.

Thanks for clarifying. I remember being interested in the T-Tap when I first saw it at BVE around 2012 or so, but the lack of reference in put me off, and regardless of that Pro Tools wouldn't work with AJA stuff. Was glad to read your review and have now bought one, so thanks.

Having tried a Blackmagic Intensity Extreme I was never happy. It would refuse to play 1080p25 and always had to be set to 1080i50, this was the same with other video software. It also caused Pro Tools to have problems with the AVE which I had never seen on my system before. Sometimes it was fussy about Thunderbolt cables too.

Blackmagic stuff has a very nice price point, but you get what you pay for.

[Frank Kruse]

Originally Posted by MIKEROPHONICS

I have abandoned my BM Decklink Studio card and gone to an AJA IOXT as my previous experience showed me that BM drivers are just plain unreliable at best.

During the time that I was having endless problems in PT11 with my BM hardware. Alan was having far less grief on the windows OS. Ok it was a different card to mine, but my ears pricked up.

Having AJA hardware compared to BM is night and day in my experience.

When the T-Tap came out I tested it to see how good an Un referenced card can be - stonkingly good is the answer. With the difference barely worth the cost.

I wanted to test the BM equivalent and try a comparison to windows as best as I can.

I am not a test lab with unlimited resources working under nitrogen atmospheres and have limited time as a busy post pro, but I make an effort to publish my findings to help others.

I started with two hypothesis;
1) BM drivers are better on windows based on anecdotal evidence from Alan
2) The Min Monitor should give reasonable non referenced results, based upon my experience with the T-Tap.
3) BM do not write good drivers and playing driver roulette is unavoidable.

The results - yes slightly limited and slightly flawed in the lack of 100 % comparison don't disprove 1), disprove 2) and seem to agree with 3). Take from them what you will.

Any tests that are not on an identical rig to yours will not be ideal for you - for that I can't help.

It Feinstein points to The latest BM driver being as bad as many of the others, which is why I bailed on them. The latest AJA driver however is absolutely fine on 10.10.3 and worked out of the box on my IOXT. Chalk and cheese.

I have not found any software that cares about my hackingtosh. It is very close to mac hardware and is chosen for the task by experts who I have confidence in. So I discount that.

The BM would not work without flashing on 10 bit full, (which is my default setting on AJA hardware) or the intermediate setting, but would only work on the draught setting.

I rarely use 24f video as I oredominately lurk in TV, so tested 25 frame. The file allows PT to output 25p or 50I from the same file with the same reference. Identical results.

I am sure the BM might work somewhere with some OS/driver settings, but for me I demonstrated that the latest BM driver is duff. This compared with my other experience makes them a poor choice imho, with maybe windows users faring better. Yours may be working for you, which is nice for you, but....

I am not trying to offend anyone, just sharing my experiences with the community and trying to make any tests as best as I can within limited time and budget.

Please feel free to conduct your own tests and publish them and we will all read them with great interest (when you have some spare time of course).

Have a nice weekend.

Thanks for the details Mike. Just to be clear: I wasn't trying to bash your test, just trying understand what you did because I'm seeing none of the things you encountered during your test. Hence my questions.

Firstly nowhere in the test you mention that it's a Hackintosh. You use the term "Mac PC" which led to my question because I simply didn't get if the term "PC" was generally meant as a word for "desktop computer" or really meant for a non-Apple machine running OS X.

The second thing is that the box you tested only worked in draft mode. This is completely not normal and I haven't seen this ever with any of the BM boxes incl. the one in your test.

In fact we only use the "full (non-10 bit)" setting on all of our 5 PT11 rigs. One of them using the miniMonitor, the others being Decklick SDIs and none of them show this behavior. Neither with 24fps, 25 nor 23.98. I have done long films all three rates over the last driver iterations with no such issues excluding the driver you used.

So it could well be a driver issue. Which led to my personal conclusion that this box with the setup you used is not really a godd combo on OS X which is indeed good info.

But it however does not mean that this box generally runs with a sync span of almost 8 PAL frames.

What you found for the windows version actually is more like what I'm seeing here under OS X but maybe I'm just lucky.

The setup where you play a 25P video at 50i measuring at 25fps is simply confusing and not further clarified in your test. You didn't mention that there was no difference between the 50i and 25P setting. Hence my question asking why this is and what the 7.something PAL frames refer to. The 50Hz frequency or the 25Hz.

Anyway, thanks for sharing the additional info.


Frank.

[Frank Kruse]

Originally Posted by tom_lowe

Thanks for clarifying. I remember being interested in the T-Tap when I first saw it at BVE around 2012 or so, but the lack of reference in put me off, and regardless of that Pro Tools wouldn't work with AJA stuff. Was glad to read your review and have now bought one, so thanks.

Having tried a Blackmagic Intensity Extreme I was never happy. It would refuse to play 1080p25 and always had to be set to 1080i50, this was the same with other video software. It also caused Pro Tools to have problems with the AVE which I had never seen on my system before. Sometimes it was fussy about Thunderbolt cables too.

Blackmagic stuff has a very nice price point, but you get what you pay for.

I've in fact just finished a 1080P/25FPS project moving between 2 protools rigs. One with a Decklink SDI the other with a 3D extreme.
Both I could switch between 50i and 25P just fine.

Frank.

[Frank Kruse]

Originally Posted by tom_lowe

Hi Frank,

I believe Mike tested the BM Mini-monitor on Alan Sallabank's PC as well as his own Pro Tools PC (running OS X) which was the exact same system he used to test the AJA T-tap. Seems a pretty fair comparison to me.

As for the lack of Ref input, the results from the AJA T-tap show this is no where near as necessary these days as it was in the past. The only real reason I can see for Ref input is if playing out to video decks.

The hardware was the same but different OS and different PT-version.

The past has shown that sometimes even moving one 0.0.X PT version up can change a lot regarding video sync. Hence my question.

In the past there have been several PT11 updates where the sync offset changed significantly after an update. For that reason I always re-check sync-offset after updating PT especially when they fixed things in the AVE like in v11.3.2.

Quoting the BM-test:

Whereas the AJA T-Tap on the same system had only an 8 ms average difference

We all know that the same computer with different OS and different PT-version can not really be called "the same system". At least that's my experience.

Frank.

[Cheesehead]

I just moved to PT 11.3.1 and BM driver 10.5 on a 6 core trashcan mac with Blackmagic Studio card and the delay required for sync in PT increased a lot, not sure of the exact figures right now but previously it was 6 quarter frames and now its a lot more.

I can't be certain, but I feel it may be to do with BM's 10.5 driver, which appears to be radically re-written from previous versions, certainly in the interface.

I also now get occasional picture tearing and nudging has become more treacle like .
I'm using Pro res proxy 1080 @ 24fps.

I need to go back to the previous BM driver and test again to be sure.

I would definitely like to try an AJA solution if it would get around these issues and get back to the snappy video performance we had in PT10.