In August, I called Dwarflab Draco a dragon hunting inside the DWARF 3’s kingdom. At the time, that headline was a bet. The only number DWARFLAB had confirmed directly was 90mm of aperture. The 340mm focal length, the f/3.8 ratio, the exposures up to 300 seconds, the cooling, the guiding was all sorted into a running list of confirmed, inferred, or reported data points. That was all there was to go on that week.
Draco has since launched. Every one of those inferred and reported numbers held. And my own Draco, the SHO edition, should arrive in November. Can’t wait and in the meantime here are some updates.
DWARFLAB supplied me with raw data from a Draco smart telescope test unit at a remote lab for independent processing. I didn’t not capture this data and I’m very grateful for having access to this data. Amongst other targets, the dataset contains a session on the Eastern Veil Nebula, Caldwell 33, a target I’ve already imaged multiple times with the DWARF 3.
This post examines and compares that Draco dataset, containing seventeen 300 second subs captured under Bortle 3 skies, against 273 60s subs I captured under Bortle 6 skies.
I also included two DWARF 3 captures of the same nebula I captured as part of my filter experiment with the DWARF 3’s built in 30nm duo band filter, and a front mounted 7nm L-Extreme duo band filter.
None of these were shot under matched sky conditions or integration times, so what follows is a first look at what each setup recovered, not a controlled head to head test.
A Nebula With History Here
The Eastern Veil, Caldwell 33, has shown up on this site before. Back in July, my DWARF 3 put 4 hours and 33 minutes into it across 273 frames on the built in Duo Band filter. It’s one of the images used in August to argue that the DWARF 3’s wide field still wins on large targets. More recently, a controlled test ran the same nebula through two different filters on the same night: the built in 30nm Duo Band against a 7nm Optolong L-eXtreme mounted externally on a bracket built from foam board, I opted for the L-Extreme as replacement for the SVBONY 7nm filter that turned out to have a wedge fault in the glass, diagnosed with a star trail test on Deneb, before it went back for a return.
That test measured both frames directly instead of eyeballing them: signal, background, and noise, region by region. The built in 30nm filter came out modestly ahead at 25 to 26 minutes: 2.39 versus 2.22 SNR on one filament, 1.39 versus 0.96 on another. The narrower filter’s noise floor really was cleaner, 40 to 45 percent lower, but the light it rejected along with the skyglow cost more signal than that noise improvement bought back. Twenty five minutes, it turned out, wasn’t enough time for a 7nm filter to pay off.
Now there’s a third pass at the same nebula, and this one isn’t a variation on the DWARF 3 at all. Different telescope, different sensor, different sky, five times the exposure per sub.
What DWARFLAB Sent
The dataset came as raw subs plus darks, flats, and bias, and a stacked fits and png processed on the device. Captured with DRACO by the DWARFLAB testing team under the stated conditions. The material was supplied by DWARFLAB for analysis. I did not physically operate the unit for this evaluation.
DWARFLAB’s own creator guide for this data makes one thing explicit: the Veil dataset came from a Draco Remote Lab unit under Bortle 3 skies, noticeably darker than an earlier Bortle 6, half moon test set they’d sent out before it. The guide asks that the two be treated as separate conditions, not a like for like comparison. I’m carrying that same warning into this post and extending it one step further. None of the three datasets below share a sky.
The two DWARF 3 frames below are mine, shot from my own Bortle 6 site. No DWARFLAB involvement there.
Three Passes at the Same Nebula
| Setup | Filter | Sub length | Sub count | Total integration | Gain | Sky | Captured by |
|---|---|---|---|---|---|---|---|
| DWARF 3, built in filter | Duo Band, ~30nm | 60s | 26 | 26 min | 90 | Bortle 6, moonless, clear | AK |
| DWARF 3, front mounted filter | Optolong L-eXtreme, 7nm, 2 inch | 60s | 25 | 25 min | 90 | Bortle 6, moonless, clear, same night | AK |
| DWARFLAB Draco, built in filter | Duo Band, 13nm | 300s | 17 | 85 min | 60 | Bortle 3, Draco Remote Lab | DWARFLAB testing team |
The two DWARF 3 rows above are the same frames measured in the filter test linked above, unchanged. The Draco row adds more than double the integration of either DWARF 3 pass, a narrower filter again, and a noticeably darker sky. Nothing here is normalized. I’m not going to pretend three sessions at three sites add up to a controlled test.
How I Processed the Draco Set
Siril, using the darks, flats, and bias DWARFLAB included:
- Naztronomy stacking script, darks/flats/bias applied
- Crop
- Gradient removal script (beta, default settings)
- Deconvolution
- Plate solve
- SPCC
- Siril noise reduction (default settings)
- StarNet star removal
- Stretch on the starless image, asinh plus GHS
- Recombine stars and starless
- Export to TIFF
- Final edit in Pixelmator Pro
The Dwarf 3 images followed a different more simpler process. I used Stellar Studio for Auto adjustments and then did final touches in Snapseed on an Apple iPad.
The Images
What 300 Second Subs Bought
Back in August, 300 seconds was the single most argued over Draco number on this site: five times the DWARF 3’s 60 second ceiling, and not yet on an actual spec sheet. It’s real now, sitting in eleven subs of actual Veil Nebula light instead of a line item in a reported features list.
The starless version I was able to produce, and I’m not a Siril expert by any means, left me speechless. The detail is just amazing and I’m looking forward what this telescope will produce in my Bortle 6 backyard this late fall and throughout the winter. What is clear to me already is that running both the Dwarf 3 and the Draco side by side promises to be productive set up to survey and record the night sky.
What This Isn’t, and What November Changes
This is a Draco Remote Lab test unit, run by DWARFLAB’s own testing team, under a sky I don’t have access to. It isn’t a retail unit in my hands yet.
That changes in November. My own Draco, SHO edition, ships then. Once it’s here, all three setups get run from the same site, the same night. The DWARF 3 has been surveying this site’s kingdom for a while now. This post is the first look at what’s about to start hunting inside of it.
Clear skies!
Frequently Asked Questions
Who captured the Draco data used in this post?
DWARFLAB’s testing team captured the raw frames on a Draco Remote Lab unit under Bortle 3 skies and supplied the raw subs and calibration frames for independent processing. The processing and final image shown here were done independently, without DWARFLAB’s involvement.
Why aren’t the DWARF 3 and Draco results directly comparable?
The three datasets come from three different nights at three different sites: a Bortle 6 site for both DWARF 3 passes and a Bortle 3 remote site for the Draco set, at different integration times. DWARFLAB’s own guidance for this data specifically asks that datasets shot under different sky conditions not be treated as a like for like comparison.
How does the Draco’s built in filter compare to the DWARF 3’s?
The DWARF 3’s built in filter is a roughly 30nm duo band. The Draco’s built in duo band filter is narrower, at 13nm, covering Ha and OIII.
Where can I see the DWARF 3 filter comparison in more detail?
A separate post on this site ran a controlled, same night test between the DWARF 3’s built in 30nm filter and a front mounted 7nm filter on this same target, with measured signal to noise numbers for both.
Leave a Reply