Burst Electronics Pixie-FS TBC review?
 

I saw an eBay listing for a Burst Electronics Pixie-FS TBC the other day and was wondering how it compares to the "later", often recommended TBCs like the AVT-8710 and TBC-1000, etc. This one is compact in size like those - is it roughly comparable or better than them?

Here's more info on the TBC from Burst Electronics' website. It retails at over $700 apparently - what makes it so different / special compared to the less expensive consumer TBCs?

2022 update:

Since this old thread has been bumped, it must be understood that prices mentioned in these 5+ year old posts is not valid, no longer accurate to the TBC market. DataVideo TBC-1000, for example, is routinely now $1500+, sometimes more than $2500 for a single unit in refurb'd condition. So a random Pixie may be worth $825 now (still doubtful), but definitely not then. These initial posts were written back when both used and new TBCs were more plentiful.

And my review of this device is in post#13: http://www.digitalfaq.com/forum/vide...html#post66599



original post:

For $825 MSRP, I could buy 2-3 DataVideo TBCs.
Even if it were decent, the eBay auction would need to around $225 to be realistically priced. Is it?

EDIT: I see the eBay listing, and it's $100 currently. It's a gamble. There are zero reviewed for this thing online, which is never good. It may just be a Cypress TBC in a custom shell.

Right now, it's an auction @ $100 with no bids and I'm the only watcher with about one day left on the listing, which is why I'm strongly considering this device.

I had a couple opportunities to get a green/black (i.e. most likely not defective) AVT-8710 for $200 shipped, but I didn't have the funds available until now. And of course the eBay listing ended and hasn't been relisted again so far...

UPDATE:

I passed on it and just won a TBC-1000 for $145 instead. Thanks for your feedback.

Based on the rather meager documentation it is an infinite window frame synchronizer, with genlock capability and some built-in proc amp capability. Can do NTSC or PAL, and can do Y/C separation from a composite input. The output format (Y/C or composite) is selected by in an internal jumper setting.

It claims to have 9-bit A/D, 4:2:2 processing, a 4-line adaptive comb filter, and offers AGC keyed to sync level, 52 dB S/N, and frequency response down 1 dB at 5.5 mHz, and is a compact package.

FWIW: it uses BNC connectors, not a cheap thing to do. Would be interesting to know more about the internals. Based on the GOOGLE hits, it is not a widely listed device.

Eh, that doesn't mean much. The Big Voodoo had all sorts of features, used BNC, and it was a paperweight.

Had hands on to examine the Burst Pixie-FS innards, but not to do electrical tests.

The guts include the following ICs:
AT17LV256 EE Programmable Serial Memory
LM39401S power regulator
TI 5150AM1 NTSC/PAL/SECAM video decoder
24LC32A Serial Electrically Erasable PROM
ATMEL ATMEGA168 Flash memory
XILINX Spartan XC2S15 field programmable gate array
AVERLOGIC AL422B-PBF 32Mb dram (4 each)
LMH6643MA dual output amp (130 MHz bandwidth)
NXP SAA7113H 9-bit video input processor (A/D)
NXP SAA 7121H digital video encoder (digital YUV to s-video and composite analog)

- The manual says you should have a vector scope and waveform monitor to use the device.
- That is, apparently no way to tell what the stored settings for contrast, brightness, saturation, hue, etc. are except by observing the waveforms.
- s-video or composite output is selected by an internal jumper. The other internal jumpers are not documented.
- s-video or composite input is selected with the Mode and Inc/Dec switches, as are proc amp, NTSC/PAL, and genlock settings settings
- It requires a 5 volt DC, 0.5 amp wall-wart to power it.
- The case is extruded aluminum ~0.06" thick
- overall about 5 1/4" x 2 3/4" x 3/4" including connectors.
- build quality look decent, a cute device, but not one for folks who like knobs, switches and readable settings.

I got the Burst Pixie-FS TBC today and opened up the hood to take some pictures, Tests will follow after I gather some tapes from the thrift store:

http://www.digitalfaq.com/forum/atta...1&d=1580785990

http://www.digitalfaq.com/forum/atta...1&d=1580785990

http://www.digitalfaq.com/forum/atta...1&d=1580785990

http://www.digitalfaq.com/forum/atta...1&d=1580785990

http://www.digitalfaq.com/forum/atta...1&d=1580786020

http://www.digitalfaq.com/forum/atta...1&d=1580786020

OOooo.... interesting. But let's see how it actually performs.

Given the above data, then your images, this may actually be a Cypress as I guessed many years ago. But the problem is then good vs. bad chips (aka green vs black AVT-8710), meaning nothing new for sale is any good. Only old models are acceptable.

We don't know if these Pixie devices are new old stock or newly made, Only tests will show how good they are. I still have the black Avttool, I want to get rid of it but I don't want to break someone's hopes, Maybe I'll list as "defective", After all that's the best word to describe those boxes anyway.

Based off chip ages, your exact Pixie unit looks like really old NOS.

The SAA chips are philips branded, which would suggest it's a bit older. Newer ones would have NXP branding instead, slightly older ones were produced by Trident before they went bust.

The SAA7113H sits between the SAA7111 used in the datavideo units and the SAA7114 that's in the cypress ones. It's also been used in a bunch of older capture cards, including the original EZCap.

The first test, total failure, This thing has a weird zebra/checkerboard pattern on almost all colors, It must be defective, Going back to the seller. Here is a screen grab:

This is literally the worst TBC I've ever used. And that's having used lots of TBCs. :huh1:

Both of my units are complete POS. Both are definitely used. One has marker scrawled on it, and the other has a broken INC selector button. Not NOS whatsoever. Used and abused.

As usual, the seller claimed "very good condition, clean, fully tested". All of that was BS. Not good condition (broken), not clean, and clearly no testing could have been performed.

- The output of the 1st unit is just random distortions.
- The 2nd has the same frame sticking problems of a black AVT-8710 -- actually, MUCH worse. The JVC menus are not just sticking, but insanely distorted, and chroma and luma have differing distortion. The distortion is similar to what you can see from an Elite BVP4+ when you crank the knobs hard left or right, and the values are out-of-bounds/illegal.

Because the POS TBC has zero reset functionality, there's no way to know where any of the values currently are, not being the original owner. The official document states to simply "used a scope". So you find yourself randomly pressing the INC/DEC (increase, decrease) buttons.

This TBC was designed by morons, and should be avoided at all costs. :2cents:

I vaguely recall testing a new Pixie at least 10 years ago, and came to a similar conclusion then. I forget what exactly was wrong with it, but I do recall rating it "DO NOT BUY", lumping it with Big Voodoo and broadcast rackmount units. Complete worthless, and ridiculously expensive (at the time, $800+ when MSRP for DataVideo was $500+), so double punishment.

I'm sure many a clueless "professional" bought these over the years, to go with their quality-losing 4:1:1 Canopus ADVC DV boxes, and low-quality 6-head VHS VCRs. :rolleyes:

After we return these, and report as damaged malfunctioning goods, I'm sure the seller will just slap them back onto Amazon, recycling the "very good, clean, tested" lie. Somebody else will be a buyer victim, but the new buyer may be clueless enough to think that's how TBCs are supposed work, so the buyer will use it to convert ruin his/her videos. Worse yet, the buyer may see the bad quality, then deride all TBCs as bad, eventually discovering the still-bad use of a Chinese HDMI converter like certain Youtube rubes claim to be "best" (but better than worst still isn't good, and certainly not "best").

Another test, another waste of time.

I'll take some test captures this week, before returning, and post those here.

Do dpalomaki's Reset instructions work?

I've already printed out the return shipping label, thank god I don't have to pay to ship it back, I will leave a nasty feedback at the product's page on amazon.
Lordsmurf, Removing the power cable resets all the values to default, Just to help you doing the tests, I know you're returning it, So you will have to set the S-Video/Composite and PAL/NTSC values every time you disconnect it from power.

I note the regulator IC, LM3940, has what appears to be a tantalum capacitor next to it (orange thing in the attached photo). I wonder if it has failed? I have had gear exhibit a high failure rate with them.

Also possibly that red goo is shorting out something, I don't see myself going into the trouble of fixing it. Maybe Lordsmurf wants to try it if he has some electronic components laying around.

Nope. Though #1 may be because the INC is broken, and #2 outputs a garbled signal.

Nope. #1 arrived with off values, and yanking power doesn't do anything. For example, #1's brightness was insanely high, and contrast was insanely low. Unplugging #1, testing #2, unplugging #2, then test #1 again showed the same adjustments.

DataVideo and Cypress remember settings session-to-session, and require manual resetting. (I actually like this, because I've always preset TBCs when selling to others, they can plug-and-play. The VCRs, by contrast, always reset, must step a person through it before 1st use.) So while it does seem to share some Cypress chips, it's not a clone.

I actually wonder if Pixie is a (poorly) reverse-engineered Cypress? :hmm:

These did not reset on the unit not giving me garbled output. No idea what the garbled unit is doing, because it's just outputting crap.

Nope. Perhaps if this was a known-good TBC, I'd be more inclined to repair it. But repairing a problem unit is just wasting time. Sort of like repairing a Funai VCR, completely pointless, worthless whether functioning or not.

I will spend some more time playing with it, but not much, more important tasks to do.

That's weird, Because my unit looses all the settings when I power it off, I set the brightness to high until everything clips off and power it off and on and the video looks normal, I did the same thing with other settings.Though I'm not sure if the default settings are the right settings.

Interesting. Proof of the warnings about buying used gear from ebay.

Have my hands on the unit I looked at 3 years ago. Powered it up and fed several test patterns to it including color bars and it appears to work. Also fed a bit of Fantasia VHS to it from an AG-1980. The default brightness and contract were off a bit (no time to adjust them yet), but the colors are correct in the vector scope. The black burst, white and gray waveforms appear to pick up a bit of noise fuzz compared to the generator output, visible on an o-scope but not obvious on the monitor I was using.

The unit is a PITA to adjust, the dials is tiny, the inc/dec buttons small and too close to other ports, and the manual is sparse at best. While most settings are stored it takes ~15 seconds for a setting to be stored. Having waveform monitor is a important requirement.

This appears to have been designed to sit in the back of a rack having been adjusted to meet the requirements of a fixed configuration modest cost studio and control room layout where one can set it to compensate for cable runs, etc., and forget it because nothing changes on a day-to-day basis. It is not especially suitable for applications where it may be necessary to tweak settings for each new tape/source several times a day

And I have no idea how well it handles sloppy sync source videos.

I know the pair of units I currently have (and will be returning for refund shortly) are worse than whatever I tested 10 years ago. I'm betting it had production changes, especially evident with the frame ghosting/sticking that matches black AVT-8710 units (but far worse).

I have no idea who the target audience for this device was. It's really not suited for pro/broadcast facilities, nor home users.

These came from Amazon, but essentially the usual "tested" and "working" eBay nonsense.

Maybe something in between, such as a small market CATV facility, or perhaps a shopping mall "rent facility time" video editing hobby shop facility such as were somewhat common ~20 years ago.

Update! :cool:

In the past month, between tasks, and especially with being a prisoner in my home (now in week 3 of self-quarantine to avoid the coronavirus pandemic, and many days of heavy rain), I've performed lots of TBC research. I took lots of notes, lots of sample images and clips, and will someday hopefully have time to share all of those in longer writings. But for now, here's the details.

I was able to acquire another set of PixieFS units. Unlike last time, these work.

latreche34 saw weird image issues: http://www.digitalfaq.com/forum/vide...html#post66589
So did I.
After experimentation, this appears to be directly related to the T-OUT-CVS jumper. And BTW, I have no idea what that jumper is supposed to do. It's undocumented. You can see that jumper in latreche34's image here (middle and right), and it needs to be flipped (middle and left). dpalomaki stated that the internal jumper selected s-video/composite output, but this jumper did not have that behavior on these units. (BTW, that likely means there are multiple revisions on the unit, though I had already surmised that from other tell-tales.) Once that is switched, the image can be quite clean and nice.

Can be.

This unit hates JVC menu, which are well known to not be 100% perfect NTSC signals. This is why the "JVC menu test" is good at finding flawed Cypress TBCs (the latter era black AVT-8710 and comparable). To the untrained eye, the Pixie appears to be a flawed Cypress as well. But it's not. Yes, it has issues with signal imperfection, and even apparent frame sticking under certain condition.

I've not detected any missing, stuck, or repeated frames as you'll see with the bad Cypress, on the actual converted footage. Only on the JVC menus, and FF/REW-while-playing operations. And that would make some sense, because it uses "green" era Cypress chips. My hypothesis is that the PixieFS is badly reverse-engineered Cypress TBC, at least at the core.

At the same time, it's not a weaker TBC like the TBC-5000, which is seemingly for the same audience (non-VHS conversion). It can be, but it's really more about the source quality and the VCR/camera model in the workflow. With some hardware, it's terrible. With others, it's quite adequate, and quite strong.

Strong, you ask? This TBC tries to do too much. It's obviously meant for non-consumer sources, not VHS or Video8/Hi8. And given the price and marketing, this makes sense. In addition to doing frame level timebase correction, it seems to attempt some sort of line correction. And it does so quite badly. So bad, in fact, that under certain conditions, it conflicts with the line TBC in the VCR, and that's what results in the frame sticking. In my ugliest line TBC tests, the Pixie actually scored well for the attempt, but the actual converted footage was still unwatchable. That same test scores AND views well ONLY with the ES10/15. The Pixie attempt scored better than the AG-1980P by itself.

With my full-tape test, I even saw small signs of temporal NR, using a Panasonic with the TBC off. Interesting.

If you
- can find a bargain on this unit ($X00s max, understanding that recommended TBCs are all now 2x+ that cost)
- have non-JVC gear like the Panasonic AG-1980P or a Hi8/D8 camera
- have retail or master (not copies) source tapes
- and only have a small collection of tapes
then this can work for you. It can be a bargain that suffices for your small project.

I still don't recommend it as a general good-for-everybody TBC. But I'd nod my head in understanding if it was used in the above scenario, similar to how I nod in understand for NTSC Mac users that use Canopus DV boxes. It's not best, but it can function in some narrow situations.

Right now, in the marketplace subforum, I'm selling both of my units.

Since there are multiple revisions of this unit, with most of them being highly suspected to use the flawed "black" Cypress chipsets, I would urge you to avoid randomly buying this from eBay or wherever. Odds are you'll get screwed with a lemon. Definitely lose time, lots of headaches/grief. Maybe lose money.

Again, not recommended. On the digitalFAQ.com class scoring, 1st/2nd/3rd/4th, this unit gets a 3rd at best (my in-hand units), and 4th at worst (the "black" chipped units).

The saddest part is that the unit is clean, the transparency is essentially perfect. The passed signal from my Panasonic AG-1980P is immaculate, and puts almost all other TBCs to shame, including the "green" Cypress and most DataVideos. And that's a real shame. Yet another analog video device that failed to live up to promises or expectations, but got so close.

The contrast/color/brightness is also nice on this pair, on par with other good Cypress.

So, there you go, a thorough review of units. :)

Thanks for the more detailed review.
The Genlock feature perhaps a bit unusual compared to others units.

FWIW: The attached graphic covering the jumper is from the Burst Pixie-FS manual, V1.02.
This implies there is at least a V1.00 and V1.01 manual as well, so that could be interpreted as indicating there are at least 2 or 3 versions of the device.

Also, I compared the board in the unit in my hands to the photos posted above. There are a number of differences including (but not limited to):
- no green jumper wire
- a 6-pin chip of some sort soldered on the center underside (perhaps a 27 mHz oscillator)
- no red 'paint' along the edge.
- the 4 pins by the top-right jumper block are not present
- an R212 "Stand-Alone Freq Adj" pot on the underside
- a surface mounted "D3" on the underside
- L5 and C58 not on underside
- Board labeled TBC8 REV E

Clearly there were significant production changes/revisions over time. Comparison of serial numbers might be interesting.

Cypress also has 0.x versions. But I don't recall ever having seen anything below a 0.9x (with x being a few as well). I remember V0.92 on a unit in years past. It was a really old model from pre-2005.

Not sure what PixieFS did. Again, to me, it appears to be reverse engineered Cypress.

I'm wondering if this exact PixieFS in my possession is using both pre-2010 and post-2010 chipsets. So it has some oddities of the "black" Cypress (AVT-8710, etc), but not the catastrophic ones that make it an unusable POS with random frame sticking and glitching mid-capture.

I had a choice to return these units, or keep them for others. I chose the latter, since TBCs are so hard to come by now. I just hope I don't regret it, want to recoup my funds. I'm not buying any PixieFS in the future. And suggest others don't gamble on it either -- outside of the pair I have available now in the marketplace, as these are thoroughly tested to actually work decently, unlike other PixieFS we've seen.

I don't know who came up with the design first, it's a pretty basic design the Datavideo TBCs are set up the same way, all of these TBCs are based around off-the-shelf video decoder/encoder chips and FPGAs (+ a ROM chip). All of them use Philips/Trident/NXP video decoder chips that are also found in a ton of other devices. The digital video conversion section of the Samsung SV-5000/7000W VCRs are actually also quite similar, using the same SAA7113 chip, but some special video conversion chip instead of a run-off-the-mill FPGA chip.

LordSmurf sold me one of his aforementioned PixieFS TBC's... and as a complete novice I'm feeling a bit lost here. This TBC appears to have a lot of functionality I will not use, without any sort of display to let you know the current settings, and I'm getting a little confused. Looking for a little help. The documentation online for the Pixie-FS equipment is... well, crap.

Looks like this TBC uses BNC ports instead of composite... despite ports that say "composite in/out". Is this correct? It would explain why I am getting no video signal when I plug female composite cables into this TBC. If this is the case... I'd imagine I need a composite or s-video converter to BNC, yes? Using composite cables, I have tried toggling between composite/s-video (mode 9) on the TBC but end up with no video regardless. I'm using the composite in/out ports of the TBC. I was under the impression that BNC ports and composite cables were the same, but I guess I am incorrect.

My VCR has S-video out... if I need to use a BNC converter, will this result in a loss of video quality? S-video to BNC converters seems to split into Chroma and Luma ports. Which ports would these go into for the inputs/outputs on this TBC? As a reminder, the unit has Composite Y, C, and Genlock input ports, and Composite Y, and Chroma outlet ports.

One more thing that's a bit more general - I have a Panasonic VCR with an internal TBC that I can toggle on/off. Once I get the PixieFS working, should I be turning off the internal TBC on the VCR? I figure running both could lead to some interference / overcorrecting.

Amen! It is meager at best.

BNC and RCA are connector types. They can be used on audio, video, RF, etc. ports. Composite is a type of signal. It typically uses RCA connectors on consumer gear, and BNC connectors on industrial/professional gear. S-VIDEO is type of signal that has also lent its name to the 4-pin mini-DIN connector and cable commonly used for it, but S-VIDEO can also use RCA or BNC connectors. (For example the Commodore Amiga monitors used RCA connectors for their S-VIDEO, composite, and audio inputs.)

The PIXIE-FS has BNC connectors instead of RCA or Mini-DIN s-video. You will probably want some BNC-to-RCA adapters, and an s-video-to-RCA wye splitter. The losses will be minimal if you make good, clean connections. Samples below (not an endorsement of these products):

https://www.monoprice.com/product?p_...SABEgJCAvD_BwE
https://www.amazon.com/CGTime-S-Vide...s&sr=1-12&th=1

The PIXIE-FS has an internal jumper setting that determines S-VIDEO or composite output. Mode #9 tells it how to interpret the input signal (composite or s-video). It is documented in the manual.

You will not need the genlock port.

In s-video in/out mode the luma signal is at the Y/Composite jack and the chroma is at the C/Chroma jacks.

Try each way, use what ever look best. It will depend on the tape you are playing.

Not covered in the manual is factory reset (e.g., to unity gain). I believe you turn the mode dial to "0" and hold both "INC" and "DEC" buttons when you apply power.

In reality you should have a waveform display/vector scope when using the PIXIE-FS so you can see what the adjustments are doing to the signal, otherwise you are kind of running blind and trusting your monitor.

For this unit, you need the Y-adapter, BNC>s-video.

I completely forgot about that. :smack:

I actually do not like the Amazon wires, and prefer shorter cheaper cables from China.
eBay search "s-video BNC Y"
For example: http://rover.ebay.com/rover/1/711-53...r/183994100248
BTW, note the delivery date. The location is fake, the seller is Chinese. So much BS on eBay!

The shielding on the Amazon cable is crap, and the longer run makes it more susceptible to RF interference. The head is also flimsy.

If you want to splurge, get Extron brand cables.
http://rover.ebay.com/rover/1/711-53...E/333206052494

Not sure what the BNC>composite adapter is for. There are better deals on eBay for BNCs. The gold is just pretty, serves no function.
I like these: http://rover.ebay.com/rover/1/711-53...m/200439230317
Out of 20 BNCs, only 1 was bad.